Reliability-based Design for Axially Loaded Driven Piles in Glacial Deposits

Abstract

Many factors and uncertainties, especially from the soil conditions, challenge the accuracy of a pile design for the geotechnical ultimate limit state (ULS) and serviceability limit state (SLS). The uncertainties can be contributed by the inconsistent material properties of glacial deposits or a lack of provisions by many design methods for the pile set-up, which is an increase in the pile capacity with time. Thus, for driven steel piles subjected to static and maintained axial loads, the goal of this research was to improve the predictability of these limit states by modifying reliability-based design (RBD) methods. A database was collected first in this thesis with a total of 120 piles subjected to static and/or Pile Driving Analyzer (PDA) tests. These tests were used to evaluate the accuracy of several existing design methods that predict the pile capacity with in-situ measurements, particularly by the standard penetration test (SPT) or piezocone penetration test (CPTu). Since existing methods commonly overestimated the capacity or exhibited significant variations in their predictions, correlations were conducted between results from the in-situ and pile tests to develop new design methods to better consider the soil content and set-up time. Many existing set-up prediction methods rely on time-consuming laboratory tests, but this study offers a practical CPTu-based approach. Afterwards, reliability analyses were performed to help designers account for the uncertainties in designs and estimate the factored resistance for the ULS and SLS. Although a few studies have previously incorporated set-up into RBDs for the ULS, very few, if any, have incorporated set-up for the SLS; thus, this thesis presents a series of resistance factors that were calibrated by Monte Carlos simulation for a range of set-up to end-of-driving resistance ratios and allowable settlements. The findings demonstrate the benefits of versatile methods that consider a range of conditions, such as set-up time, soil classification, and pile geometry. In addition, economic benefits may be received for designs by using a greater factored resistance at a set-up time compared to the end-of-pile-driving condition. Overall, the findings can assist practitioners to mitigate geotechnical risks and deliver more economic and reliable pile designs.